Back in 2008 before starting with high voltage, I have made numerous experiments with flyback transformers and with simple single-transistor drivers, most of these were undocumented. However, this was one of my very first high voltage circuits and my first -electronic- circuits in general, which were marked with some success. The resulting simplicity of this circuit pardons the low output efficiency…
This circuit, also incorrectly known as the ’2N3055 flyback driver‘, bears some resemblance to the battery inverter/continuity tester which is published here. It is the same topology with almost the same “efficiency”, and it’s actually a nice example what happens when a primitive circuit like this is used with slightly bigger input powers . High voltage enthusiasts surely know what I am talking about – this circuit is woefully unstable, inefficient, heats up immensely and it is seriously prone to failure when overdriven over 12 volts. The main reason is the lack of protection circuits, no stabilization, no control over the frequency, feedback and a lack of methods to improve this flyback topology’s efficiency. However, this circuit still serves its purpose: it is a gateway of every beginning high voltage experimenter to the world of arcs, sparks and alike (since it is simple to construct); plus the low output power makes an accidental contact with flyback’s energized output non-lethal. It will slap and kick like hell and maybe some minor burns might occur, however it teaches the experimenter to have respect before moving to more efficient and powerful drivers. And the output can also light up a candle easily (well, if the flame is not extinguished by the anode’s ionic wind, like in the video, unfortunately).
The reason why I’ve put the 2N3055 driver in quotation marks is simple. This circuit isn’t precisely the “2n3055 driver” you have seen on other websites, as the two high-wattage resistors used in those schematics are replaced here with a single one – you’ve guessed it right, with a classic incandescent lightbulb. Plus, almost every new 2N3055 you will come across at auction sites and e-shops nowadays will be counterfeit and made somewhere in the “Third World”, thus they are seriously underrated in maximum ratings, especially current ratings and breakdown voltage. For this reason I have used more sturdy bipolar transistors made by the former Czechoslovak TESLA, most notably the KD5xx and KD6xx series, which could take at least 2 times more abuse than the original 2N3055 made by the U.S. RCA company… However, let’s move on!
Well, after fiddling with ZVS drivers and similar halfbridge circuits I have returned back to the original 2N3055 driver approach. Then Jan Martiš, a Czech experimenter, came up with an idea to increase the efficiency of these flyback topologies: to add a resonant capacitor to the primary winding and tuning it afterwards to get the best output.
This circuit will remind you of the “international” 2N3055 driver you are familiar with. The principle is still the same, it is self-oscillating and has a feedback winding intact, but notice the resonant capacitor in the primary which really makes wonders. That fast diode is there to protect the bipolar transistor from excess voltage spikes. And that’s all, folks!
By tuning the resonant capacitor and the input voltage, you’re likely to get thinner arcs and bigger output voltage, or fat arcs and a smaller voltage output, or a compromise, or it will fail to oscillate. The last one happens when the primary/feedback/secondary windings are in incorrect polarity or phasing, or if the input voltage is too low (18-24 volts is recommended for this circuit). Likewise, the NPN must be heatsinked like in the first unmodified driver!
This is what I’ve got with a 330n resonant cap and a soft 40 volt power supply (a “soft supply” has a huge voltage drop when loaded,in my case I used a 40 watt transformer which couldn’t supply enough current to make “fat arcs” in layman’s terms, but because of that voltage drop it was still safe for the transistor to operate):